New Questionnaire Aims to Measure How Staff Judge Mechanical Restraint in Psychiatric Care

Mechanical restraint is one of the most drastic interventions available in acute mental health care. When a patient poses an immediate danger to themselves or others, staff on psychiatric wards may secure the person’s limbs or torso to a bed with belts, a measure that is legally regulated, ethically fraught and physically and psychologically risky. The practice can cause unexpected harm, ranging from physical injury to lasting mental trauma, which is why clinical guidelines demand continuous assessment and decision-making before, during and after every restraint episode. Yet until now, researchers have had remarkably few validated instruments to measure how the health care professionals responsible for those assessments actually perceive and carry out this demanding part of their work.

A pilot study published in BMC Psychiatry by researchers at Molde University College in Norway set out to close that gap. The team, led by Liv Bachmann together with Cecilie K. Utheim Grønvik, Atle Ødegård and Ingunn Mundal, developed the Mechanical Restraint Questionnaire, or MR-Q, and subjected it to its first round of psychometric testing. The instrument is a self-report questionnaire designed to assess nurses’ and ward staff members’ perceptions of the assessments related to the use of mechanical restraint, and its debut represents a small but potentially consequential step toward making restraint practice measurable, comparable and ultimately improvable.

The construction of the MR-Q followed a deliberately multi-source approach to item development. Rather than relying on a single theoretical framework, the authors drew on individual interviews with clinicians, an existing clinical guideline, and a review of the literature to generate candidate items. This triangulation matters because restraint assessment is not a single act but a chain of judgments: deciding whether restraint is necessary, monitoring the patient while it is in place, and evaluating the aftermath. The final version of the questionnaire contained 61 items, of which 26 addressed the practice environment, meaning the organizational and contextual conditions under which restraint decisions are made, and 35 addressed the caring process, the clinical and relational work that surrounds each restraint episode.

Data were collected in a cross-sectional design from 69 nurses and ward staff working on acute mental health wards in Norway. That sample size is modest, and the authors are explicit that this is a pilot study, but it was sufficient to begin probing the questionnaire’s underlying structure using standard techniques from psychometrics. The analysis pipeline combined parallel analysis, visual inspection of the scree plot, and exploratory factor analysis, the workhorse method for discovering how many latent dimensions a set of items actually taps. Parallel analysis, which compares observed eigenvalues against those generated by random data, suggested a possible six-factor solution, and the exploratory factor analysis ultimately extracted two dimensions with three factors in each.

The resulting structure has an intuitive logic that mirrors the questionnaire’s design. Three of the six factors covered the practice environment and three covered the caring process, exactly the two domains the item pool had been built around. In terms of explained variance, the three practice-environment factors together accounted for 69.3 percent of the total variance in that domain, while the three caring-process factors accounted for 52.0 percent of the variance in theirs. For a first-pass instrument measuring complex professional perceptions, those figures indicate that the items cluster into coherent dimensions rather than scattering randomly, which is precisely what construct validation is supposed to reveal at this stage.

Beyond the factor structure, the team assessed the questionnaire through face validity, content validity, construct validity and reliability testing of the identified subscales. Reliability, typically evaluated through internal consistency measures that gauge how well items within a subscale correlate with one another, came out in the acceptable to high range. In practical terms, this means that the subscales behave as unified scales rather than loose collections of loosely related questions, a prerequisite for anyone hoping to use the MR-Q to compare wards, track changes over time or evaluate interventions aimed at reducing restraint use.

Why does a measurement instrument for staff perceptions warrant this much attention? The answer lies in the peculiar epistemic position of restraint in psychiatry. Because the intervention happens behind closed ward doors, often in emergencies, much of what is known about how restraints are decided, applied and reviewed depends on the recollections and judgments of the professionals involved. Patient perspectives are increasingly collected, but the staff side of the encounter has been documented mostly through audits and incident reports, which capture events rather than the perceptions and assessments that shape them. A validated questionnaire turns that tacit clinical territory into data that can be quantified, benchmarked and studied statistically.

The potential applications follow from that. If the MR-Q performs well in larger samples, hospital managers could use it to map how assessment practices vary across wards, identifying units where staff feel less supported by the practice environment or less confident in the caring process. Researchers could use it as an outcome measure in studies testing whether specific interventions, such as debriefing routines, staffing changes or training programs, actually change how restraint-related assessments are perceived and performed. Regulators and quality-improvement programs, which in many countries are under pressure to reduce coercive measures, would gain a standardized way to monitor the conditions under which restraint is used rather than merely counting how often it happens.

The authors are careful about the limits of what this pilot can claim. Sixty-nine respondents is a small sample for factor-analytic work, and the participants came from acute mental health wards in a single country, raising questions about whether the six-factor structure would replicate in more heterogeneous professional groups and health systems. The team states that some improvements to the questionnaire are needed and that further investigation of the reliability and validity in a larger and more heterogeneous sample is required. In psychometric practice, the natural next step after exploratory factor analysis is confirmatory factor analysis on an independent sample, a test the present study did not include, and the abbreviation list of the paper itself signals that confirmatory work is on the research agenda.

Even with those caveats, the study fills a genuine void. Coercive measures in psychiatry are under intensifying scrutiny worldwide, with policy frameworks pushing toward prevention and, where restraint is unavoidable, toward rigorous documentation and review. Tools like the MR-Q give that agenda an empirical handle on the professional side of the equation, measuring not just whether restraints happen but how the people who carry them out perceive the assessments that are supposed to keep patients safe. As the questionnaire moves from pilot testing toward validation in larger and more diverse samples, it may become a standard instrument in a field that has long lacked one, turning a hidden dimension of psychiatric practice into something that can finally be seen, compared and improved.

Subject of Research: Development and psychometric testing of a questionnaire measuring healthcare professionals' perceptions of mechanical restraint assessment in acute mental health care

Article Title: Development and psychometric testing of the Mechanical Restraint Questionnaire (MR-Q): a pilot study

Article References: Bachmann, L., Utheim Grønvik, C. K., Ødegård, A., & Mundal, I. (2026). Development and psychometric testing of the Mechanical Restraint Questionnaire (MR-Q): a pilot study. BMC Psychiatry. https://doi.org/10.1186/s12888-026-08691-w

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08691-w

Keywords: mechanical restraint, psychiatry, psychometrics, questionnaire development, exploratory factor analysis, nurses, acute mental health care, internal consistency, pilot study, instrument validation, patient safety, Norway